Remove anonymized dataset todo from deduplicator code

The deduplicator still contains a TODO about writing out the anonymized dataset, but as discussed in #23, but this is actually already done in !31 (merged) - anonymized and full dataset are written out concurrently as different Kafka topics.